Estimating the prevalence of sensitive attribute with optional unrelated question randomized response models under simple and stratied random sampling

This study proposes optional Randomized Response Technique (RRT) models in a binary response situation. The application of the proposed optional RRT models under stratication is also explored. Gupta et al. [1] introduced an ingenious idea of optional RRT model in which a question may appear sensitive to one respondent while being not sensitive to another. This study focuses on estimating , the prevalence of sensitive attribute, !, the sensitivity level of the underlying sensitive question when the proportion of unrelated innocuous attribute x is unknown. A new multi-question approach is proposed for estimation of parameters (; !). A comparison between the proposed optional RRT models and the corresponding full RRT models is carried out numerically under simple and stratied random sampling.